UN High Commissioner for Human Rights Volker Türk has called for urgent international action to establish stronger safeguards, independent oversight and clear limits on advanced artificial intelligence, warning that increasingly powerful AI systems could pose an “existential risk to humanity.”

Addressing the UN Human Rights Council on Monday, Türk warned that the rapid development of artificial intelligence was creating “unfamiliar, even unprecedented” threats to human rights and urged governments and technology companies to act before the risks become impossible to control.

“The need for AI governance is widely acknowledged, but where is the action?” Türk asked, warning that continued delays could primarily benefit powerful technology companies, their owners and those enabling the rapid expansion of AI capabilities.

Türk expressed concern over the enormous influence held by a relatively small number of individuals and companies driving the development of advanced AI.

He warned that a handful of technology leaders now wield extraordinary power over systems with increasingly vast computing capabilities, raising fundamental questions about accountability, transparency and democratic oversight.

The UN rights chief also criticized the extensive collection and use of personal data by AI systems, arguing that claims of technological freedom can sometimes conceal what amounts to the commercial exploitation of people's information.

Türk highlighted growing concerns surrounding AI agents that are capable of performing increasingly complex tasks with limited human intervention.

He pointed specifically to scenarios in which advanced AI systems could escape controlled testing environments or attempt to prevent developers from shutting them down.

“AI that escapes its testing environment or blackmails developers to prevent itself from being turned off, is AI that is too powerful,” Türk warned.

His comments followed reports of incidents involving AI agents during testing. In July, reports emerged that OpenAI agents had executed code on servers operated by Hugging Face, a platform widely used by developers to share AI software. OpenAI later said its agents had executed code on dozens of Hugging Face servers and gained full access to one of them.

Other major technology companies, including Anthropic and Meta, have also reported incidents involving AI agents during testing, adding to concerns over the behavior and security of increasingly autonomous systems.

Türk said he would write to AI companies in the coming days, urging them to take immediate steps to reduce risks within their control.

He called on countries that host AI infrastructure and those involved in its supply chains to work together to establish internationally agreed red lines for the development and deployment of advanced AI.

He also called for independent verification of AI safety measures and stronger cooperation within the technology industry on cybersecurity.

But Türk stressed that the risks associated with AI extend beyond technical failures. He warned that the technology could have far-reaching consequences for employment, democracy, human rights and the environment.

“I share the concerns of industry insiders that advanced AI could pose an existential risk to humanity,” he said, calling for “cast-iron guarantees” to ensure the safety and security of AI systems before it is too late.

Türk’s warning comes as governments and international institutions grapple with how to regulate rapidly advancing AI technologies.

The European Union’s AI Act has established a comprehensive regulatory framework based on the level of risk posed by different AI applications. High-risk systems face strict requirements, while certain uses considered unacceptable threats to safety or fundamental rights are prohibited.

The legislation restricts practices such as AI systems designed to manipulate or exploit people, social scoring, certain forms of biometric profiling and some uses of facial-recognition technology.

It also places restrictions on emotion-recognition systems in workplaces and schools and limits real-time biometric identification by law enforcement in public spaces, subject to specific exceptions.

As artificial intelligence continues to advance at unprecedented speed, Türk’s intervention adds to growing international calls for stronger oversight, greater transparency and clear accountability—particularly for systems capable of acting with increasing autonomy and operating beyond traditional forms of human supervision. (ILKHA)